0

I'm new to R and I need advice on dealing with this problem:

I have 2 tables. The start of the tables are shown below:

Table 1: 

   SNP              Gene      Pval  Best_SNP    Best_Pval
rs2932538   ENSG00000007341 5.6007
rs10488631  ENSG00000064419 7.7461
rs12537284  ENSG00000064419 4.5544
rs3764650   ENSG00000064666 12.3401
rs10479002  ENSG00000072682 5.0141
rs6704644   ENSG00000072682 6.2306
rs2900211   ENSG00000072682 9.9022

Table 2:

Best_SNP          Gene      Best_Pval   
rs9028922   ENSG00000007341 10.7892
rs8233293   ENSG00000064666 89.342
rs3234432   ENSG00000072682 32.321
rs2892334   ENSG00000064419 43.235

Table 1 contains the entire list of SNPs for each gene. Table 2 contains the best SNP and the corresponding best Pval for each gene that appears in Table 1.

I want to do the following: match each Gene from Table 1 to Table 2 and then copy the Best_SNP and Best_Pval from Table 2 and paste them in under Best_SNP and Best_Pval columns in Table 1 for that Gene. The tricky part is that in Table 1, each gene is repeated for a random, different number of rows. For example, the second gene ENSG00000064419 repeats for 2 rows and ENSG00000072682 repeats for 3 rows. So the code needs to filter through the names of the genes, and only copy down the Best_SNP and Best_Pval once for the same gene.

So for gene ENSG00000072682, out of the 3 rows, only the first row that appears to contain the gene needs to have the Best_SNP and Best_Pval columns filled in. I don't want the rest of the 2 repeated rows to also have the columns Best_SNP and Best_Pval filled in. It'll be easier to see where each gene starts and ends that way.

Brian Tompsett - 汤莱恩
  • 5,753
  • 72
  • 57
  • 129
zfz
  • 153
  • 3
  • 16
  • that's the kind of `vlookup` function in Excel. You can see [here](http://stackoverflow.com/questions/15303283/how-to-do-vlookup-and-fill-down-like-in-excel-in-r) for doing that in R. – Metrics Jul 01 '13 at 15:15

1 Answers1

0

If I understand the question correctly, this is the solution:

x <- structure(list(SNP = c("rs2932538",  "rs10488631",  "rs12537284", "rs3764650",  
"rs10479002",  "rs6704644",  "rs2900211"), Gene = c("ENSG00000007341", "ENSG00000064419", 
"ENSG00000064419", "ENSG00000064666", "ENSG00000072682", "ENSG00000072682","ENSG00000072682"),
Pval= c(5.6007, 7.7461, 4.5544, 12.3401, 5.0141, 6.2306, 9.9022)), row.names= c(NA, 7L), class = "data.frame")

x
         SNP            Gene    Pval
1  rs2932538 ENSG00000007341  5.6007
2 rs10488631 ENSG00000064419  7.7461
3 rs12537284 ENSG00000064419  4.5544
4  rs3764650 ENSG00000064666 12.3401
5 rs10479002 ENSG00000072682  5.0141
6  rs6704644 ENSG00000072682  6.2306
7  rs2900211 ENSG00000072682  9.9022

x1 <- x[!(duplicated(x$Gene) | duplicated(x$Gene, fromLast = FALSE)), ]

x1
         SNP            Gene    Pval
1  rs2932538 ENSG00000007341  5.6007
2 rs10488631 ENSG00000064419  7.7461
4  rs3764650 ENSG00000064666 12.3401
5 rs10479002 ENSG00000072682  5.0141

y <- structure(list(Best_SNP = c("rs9028922", "rs8233293", "rs3234432", "rs2892334"), Gene =  c("ENSG00000007341", "ENSG00000064666",  
"ENSG00000072682",  "ENSG00000064419" ),
Best_Pval= c(10.7892, 89.342, 32.321, 43.235)), row.names= c(NA, 4L), class = "data.frame")

y
   Best_SNP            Gene Best_Pval
1 rs9028922 ENSG00000007341   10.7892
2 rs8233293 ENSG00000064666   89.3420
3 rs3234432 ENSG00000072682   32.3210
4 rs2892334 ENSG00000064419   43.2350


merge(x1, y, by="Gene", all= FALSE)

             Gene        SNP    Pval  Best_SNP Best_Pval
1 ENSG00000007341  rs2932538  5.6007 rs9028922   10.7892
2 ENSG00000064419 rs10488631  7.7461 rs2892334   43.2350
3 ENSG00000064666  rs3764650 12.3401 rs8233293   89.3420
4 ENSG00000072682 rs10479002  5.0141 rs3234432   32.3210
S Das
  • 3,291
  • 6
  • 26
  • 41
  • hi thanks. this code seems to delete the rows where the same gene is repeated. but is there a way i can keep ALL the rows in Table 1; and if a gene is repeated, then I just want to copy the Best_SNP and Best_Pval for the **first** gene in the first row (out of all the rows where the gene repeats one row after another). – zfz Jul 02 '13 at 09:08